POS & Tech Stack

Toast wins for full-service with more than 6 tables because handhelds, KDS routing, and coursing are first-party and the menu engine handles modifier-heavy NYC menus without breaking. Square is the right call only if you are a counter-service shop under about $1M in annual sales — its $0 monthly Free plan beats Toast's $69/mo Core plan for that footprint. Toast Starter is technically $0/mo but locks you into 2.99% + $0.15 card-present rates with no negotiation; Toast Core at $69/mo lets you negotiate down to roughly 2.49% + $0.15 once you clear $50K/mo in card volume. For 100-seat NYC dining rooms, the Toast handheld at $50/mo per device pays for itself in 3-4 turned tables a night. Pick Square if simplicity and same-day signup matter more than coursing; pick Toast if your check average is over $40 and you have a real kitchen.

The current NYC craft cocktail default is Toast for any bar doing over $30K/wk in liquor sales, with Square as the budget pick under that threshold. Lightspeed Restaurant K-Series (the old iKentoo) holds maybe 10-15% of the high-volume Manhattan bar market — Death & Co, Attaboy, and a chunk of LDV Hospitality run Lightspeed because the tab management and split-check logic is faster than Toast's. Aloha NCR still runs at legacy hotel bars (Bemelmans, King Cole) but nobody picks it new — installs are quoted at $15K+ before hardware. For a new neighborhood bar, budget $69/mo for Toast Core plus 2 handhelds at $50/mo each plus one KDS at $25/mo, all-in roughly $200/mo software before processing.

If your hotel runs Opera Cloud or Mews PMS, your realistic options are Simphony (Oracle/Micros), Agilysys InfoGenesis, or Lightspeed K-Series — these are the three POS systems with certified two-way Opera interfaces that handle room-charge posting cleanly. Toast does have an Opera connector via OHIP launched in late 2024 but it is still considered beta at most NYC luxury properties; the Mark, the Carlyle, and Park Lane all stayed on Simphony through 2026 because of room-charge reliability. Budget $8-15K for the Opera interface license plus $200-400/month per Simphony workstation. If your hotel restaurant is a separate concept with its own check average and tax ID (Locanda Verde at the Greenwich, Crown Shy at 70 Pine), you can skip the PMS integration entirely and run Toast standalone.

Your bank pitches Clover because Fiserv owns Clover and pays the bank a referral kicker — typically 25-50bps of your card volume forever. The hardware looks slick but the lock-in is real: most Clover deals come with a 3-year processor agreement at 2.6% + $0.10 plus a $30/mo gateway fee, and the early-termination fee runs $295-495 per terminal. For a single-location NYC quick-service shop that processes under $500K/yr Clover is fine and you can negotiate the rate down with leverage; for full-service you will outgrow the menu modifier engine inside 12 months. The single biggest red flag: ask for the actual processor name (TSYS, First Data, Fiserv) and the buy rate, not just the qualified rate — Clover quotes routinely tier 40% of transactions into mid- and non-qualified buckets at 3.5%+.

Toast, Lightspeed, and Square for Restaurants all handle multi-unit through a single back-office, but the menu sync and reporting break differently. Toast's multi-location module is included on Core+ ($165/mo per location) and handles centralized menu push to 50+ units without choking — Sweetgreen and CAVA both run their full chains on Toast. Lightspeed's Hospitality Hub is cleaner for franchisor/franchisee splits where each unit has its own books. If you are headed past 10 units, look hard at Olo as your ordering layer (since 2005, $850M+ revenue, runs Shake Shack and Sweetgreen ordering) sitting on top of either POS. Skip Square once you cross 5 units — the consolidated reporting falls apart and you will end up exporting CSVs to Google Sheets every Monday.

Yes, but you almost certainly should not. NCR Voyix still sells Aloha Cloud and Oracle still sells Micros Simphony, but both are now positioned as enterprise-only — minimum spend on a new Aloha install runs $20-30K plus hardware and a 3-year managed services contract at $300-500/mo per terminal. Both are objectively more reliable than Toast or Square in pure uptime terms (Simphony measured at 99.99% in 2025), which is why luxury hotels and casinos still buy them. For a standalone NYC restaurant under 200 seats, the implementation cost and 6-9 month deployment window kill the deal — pick Toast or Lightspeed instead and re-evaluate at unit 10. The exception: if you are opening inside a Marriott or Hilton property, the brand standard may force you onto Simphony or InfoGenesis whether you like it or not.

Square for Restaurants Free at $0/mo plus a single Square Register at $799 one-time gets you to opening day for under $1,500 hardware out the door. At $1M/yr in card volume you'll pay roughly $26K/yr in Square processing fees at 2.6% + $0.10 — that's the budget number to compare every alternative against. Toast Starter Kit competes hard here at $0/mo with two terminals included, but locks you into Toast Payments at 2.99% + $0.15 which costs you about $3-4K/yr more than Square at this volume. The unsexy truth: at counter-service scale the POS choice matters less than your online ordering integration, because 40-60% of your revenue will come through your website and delivery apps. Optimize for ordering, not POS.

Resy in NYC, full stop, for any concept that hopes to be on the Best New list — Resy owns the demand-side mindshare in Manhattan and Brooklyn and the Amex-funded marketing engine drives 20-40% of cover volume at a hot opening. Resy charges a flat $249/mo (Basic) up to $899/mo (Pro) with no per-cover fee, vs OpenTable at $39/mo platform plus $1.50 per network cover and $0.25 per direct cover. For a 60-seat restaurant doing 3,000 covers/month, OpenTable runs $4,500-6,000/mo all-in vs Resy at $899/mo flat — Resy is cheaper above ~600 monthly network covers. The exception: outer-borough family neighborhood spots and any restaurant where SEO + Google match matters more than tastemaker buzz still benefits from OpenTable's network. Most Manhattan operators run Resy primary plus a Google Reserve listing and skip OpenTable entirely.

SevenRooms is the right pick when you need a real CRM layer — guest tags, spend history, allergy notes, VIP lists synced across multiple venues — not just a reservation book. It is the canonical tool for hotel F&B (Soho House, Tao Group, LDV, Catch, Major Food Group on the special-event side) at $499-1,499/mo per venue depending on modules. SevenRooms was acquired by Bain Capital in 2023 and bought by DoorDash in 2025 ($1.2B announced May 2024, closed early 2025), so the roadmap now leans into delivery-data integration. Pick SevenRooms over Resy if you have at least 2 venues, do meaningful private-event business, or need to pass a guest's $400 lifetime spend at the Tribeca location to the host at the new Williamsburg outpost. Pick Resy if you are a single venue and want to skip the 6-week implementation.

Yes — Tock is built for prepaid ticketed reservations and is the default for any NYC tasting-menu over $150/pp. Eleven Madison Park, Atomix, Le Bernardin, Per Se, and Crown Shy all run Tock; the prepay model crushes no-show rates from 12-18% (industry baseline) to under 2%. Tock pricing runs $199/mo Basic, $499/mo Plus, or $899/mo Pro, plus 2.5% on prepaid ticket sales — meaningful but worth it on a $295 menu where one no-show costs you $295 cash. Tock was sold by Squarespace to American Express in March 2024 for ~$400M and is now operated alongside Resy under the Amex umbrella, so expect deeper Resy-Tock cross-promotion through 2026-27. If your check average is under $100 and you don't take prepayment, Resy is still the right tool.

Yelp Guest Manager (the rebrand of Yelp Reservations + NoWait) runs $249/mo flat with no per-cover fee and is a defensible budget pick for a neighborhood spot that already gets meaningful traffic from Yelp organic. The killer feature is the integrated Yelp waitlist — diners can join from the Yelp app while standing on the sidewalk, which moves the needle for high-walk-in concepts in places like East Village, LES, and Williamsburg. The catch: Yelp's diner-side reservation discovery is dwarfed by Resy and Google in NYC, so you are using it as a host-stand tool, not a demand-generation tool. Most operators who run it pair it with a Google Reserve listing and drop OpenTable to cut $300-500/mo.

Resy lets you set a credit-card-hold no-show fee of $25-100/pp, with the most common Manhattan default at $25/pp for parties of 4 or fewer and $50/pp for 5+. To enable it without crushing booking volume, gate the hold to (1) parties of 5+, (2) Friday/Saturday only, and (3) any reservation made within 48 hours of seating — that captures roughly 70% of your no-show risk while leaving 70% of weeknight bookings frictionless. You must give at least 24 hours' notice via email/SMS that the fee will be charged, and Resy auto-handles the reminder. Track your no-show rate weekly: NYC FSR baseline is 8-15%; if a $25 fee drops you under 5% it is paying for itself many times over given a $75 NYC check average.

Yes — turn it on day one because it costs nothing extra and adds the 'Reserve a table' button to your Google Business Profile, which is the single highest-converting reservation surface on the internet for a NYC restaurant. Google Reserve doesn't run reservations itself; it routes the booking to whichever certified partner you use (Resy, OpenTable, SevenRooms, Tock all qualify). Resy and OpenTable both push to Google Reserve automatically once you enable the Google integration in their back office — usually a single toggle. Operators who turn this on report 10-25% lift in direct (non-network-fee) reservations within 30 days. The only setup work: claim and verify your Google Business Profile, set hours correctly, and confirm the 'Reserve' button appears on mobile search.

Yelp Waitlist (free with Guest Manager) and Resy Waitlist (included in any Resy plan) are the two real options, and Yelp wins for pure walk-in concepts because diners already have the Yelp app open while looking for a place to eat. Both let guests join from a phone, get an SMS when their table is ready, and walk back without giving up their spot — the operational lift is real. NYC walk-in spots like Lucali, Joe's Pizza, and Di Fara have historically refused all waitlist tech; for everyone else, expect 30-45% of weekend walk-ins to use the waitlist app once you signal it on the door. Budget zero incremental cost if you already run Yelp Guest Manager; otherwise $249/mo for the Yelp standalone waitlist module.

Run KDS unless you are a 25-seat single-station concept where the chef can see every ticket from one spot. The throughput math is decisive: KDS cuts ticket times by 15-25% in busy services and eliminates the legibility, lost-ticket, and modifier-miss errors that drive most kitchen comps. Toast KDS runs $25/mo per screen with the Core plan and includes routing logic that splits a single order across cold/hot/grill/expo without the expo doing manual sorting. For a 60-seat NYC restaurant doing 200 covers a night, two KDS screens (line + expo) at $50/mo total pays for itself in ~3 fewer comps a week. The only counter-argument is steakhouse fine dining where chefs prefer the tactile chit; even then most have moved to KDS plus printed expo chits as a hybrid.

Map each menu item to one or more stations (cold, hot, fry, grill, raw, dessert, expo) at menu-build time, not as an afterthought — bad routing is the #1 reason a $25/mo KDS feels broken. Standard NYC FSR setup runs 4 screens: cold app station, hot line, expo, and bar (drinks). Set coursing rules so apps fire on order and entrees fire on bump-from-cold-station, not on a manual fire button — this is where Toast's KDS beats most competitors. Build at least 3 colors of timer alerts: green under 6 minutes, yellow 6-9, red 9+, and have your sous-chef walk the line every 30 minutes during service to clear stale tickets. For a steakhouse or any concept where temp callouts matter, also enable the 'temp' modifier-display flag so 'medium-rare' shows as a 30pt tag on the screen.

Buy 22-inch commercial-grade IP54-rated touchscreens — the QSR Automations ConnectSmart and Toast KDS Screen at ~$799-1,200 per unit are the two NYC standards, both rated for kitchen heat and grease. Mount at eye level (60-66 inches from floor to screen center) directly above each station, with the bottom of the screen at least 8 inches above any heat source per NYC Health Code grease-management practice. Run Cat6 ethernet, not WiFi, to every KDS — kitchen WiFi gets killed by stainless steel and microwaves and you cannot afford a 30-second screen freeze at 8pm Saturday. Budget $200-400 per screen for stainless mount arms (Ergotron HX or Innovative 9136) and another $100 for a sealed touchscreen hood if your screen sits near a flat-top.

Touchscreens work fine for cold/expo stations where hands are clean; physical bump bars are worth the $250-450 each for hot line stations where cooks have hands covered in flour, oil, or marinade and shouldn't be smearing it on a touchscreen 200 times a service. The Logic Controls KB1700 and Toast Bump Bar are the two common units, both wire-mounted next to the screen. The unsung benefit: a bump bar lets a cook clear a ticket without taking eyes off the pass, which protects pacing on a busy Saturday. Skip the bump bar at expo because the expo is doing the routing and visual confirmation, and skip at fry/grill if your cooks wear gloves. Single-screen bars and small kitchens don't need them.

Route delivery orders (DoorDash, UberEats, Grubhub) to a dedicated 'delivery' KDS screen separated from the dine-in line, and stagger the cook-fire by 8-12 minutes after acceptance so the driver arrives just as food hits the pass. Toast, Olo, and Otter (the third-party order aggregator) all let you set a per-channel quote-time padding — bump dine-in to 20 min and delivery to 35-40 min during 7-9pm rush so the kitchen can prioritize dine-in pacing. The other lever is a per-channel pause: in Otter or Toast, pause UberEats and Grubhub at 8pm if your line is in the weeds, and re-open at 9. Operators who don't do this routinely see delivery food sitting 6-10 minutes at the pass during rush, which is what generates 1-star reviews and refund requests.

Industry NYC FSR target is 15-18 minutes ticket-to-pass for entrees and 6-9 minutes for apps, measured from order-fire to expo-bump. QSR target is dramatically tighter: 4-6 minutes total. The KDS dashboard in Toast/Lightspeed/Aloha will show you a rolling average — set the line cook's red-alert at 12 minutes for apps and 22 for entrees, and pull the ticket-time report every Monday morning to find your bottleneck stations. The other number to watch: re-fire rate (orders sent back from expo for being wrong or cold), which should run under 3% of tickets — anything over 5% means your routing or station mise is broken. Don't let your chef tell you ticket time doesn't matter; it correlates 1:1 with table turn rate, which is the most direct lever on your nightly revenue.

An all-day display shows the running total of every active item across all open tickets — '12 ribeye, 8 branzino, 6 burger' instead of ticket-by-ticket — which lets the line plan firing and bulk-batch where possible. Yes, you need it on any line doing more than 80 covers a service; it's a single toggle in Toast KDS or QSR Automations and runs on the same screen the chef or sous-chef is already watching. The payoff is biggest at protein stations where a cook can pull 12 steaks at once for a 7:30 push instead of pulling them in 3 waves. Counter-service and pizza shops benefit even more because the all-day shows you when to start a fresh batch of dough or fries before you run out mid-service.

7shifts is the NYC default for restaurants under 50 employees per location — pricing starts at $34.99/mo per location (Entrée plan) and goes to $76.99/mo (The Works), and the Toast integration pulls actual sales into the labor forecast in near real-time. HotSchedules (now Fourth) is the legacy enterprise pick used by Union Square Hospitality, Major Food Group, and most multi-unit hotel F&B operators because it handles 200+ employee org charts and union pay rules. Sling (owned by Toast since 2018) is included free with Toast Payroll which makes it the no-brainer for Toast-Payroll customers. Skip Homebase for full-service restaurants — it works for retail and small QSR but the FOH/BOH role split and tip pool tracking is weak.

Yes, in practice — any scheduling tool worth running needs to pull your actual clock-in/clock-out from the POS terminal so your labor cost vs sales calculation is honest. 7shifts integrates with Toast, Square, Clover, Lightspeed, and Aloha out of the box; HotSchedules integrates with Toast, Aloha, and Micros. The integration runs every 15-60 minutes and pulls (1) employee hours, (2) sales by hour, (3) tips for the period — letting you see real-time labor as a % of sales, which is the single most important number on a manager's dashboard. NYC labor cost target is 28-34% for FSR, 22-28% for QSR; if you're above 36% on a Tuesday lunch, your scheduling is wrong and the dashboard should be screaming.

If you operate a fast-food chain with 30+ locations nationally (NYC Admin Code §20-1201 et seq.), you must give 14 days' written notice of schedules and pay a $10-75 schedule-change premium for any change inside that window. 7shifts, HotSchedules, and Sling all have a 'Fair Workweek' module that auto-calculates the premiums, but you have to enable it and tag your locations as covered. Full-service restaurants in NYC are NOT currently covered, but Council bills to expand coverage have been introduced in every session since 2022 — assume coverage is coming and pick a tool that supports it. The tool also needs to keep schedules and the schedule-change log for 3 years per the law; cloud tools handle that automatically, paper schedules don't.

7shifts Tip Pooling is included on the $76.99/mo The Works plan and handles points-based, hours-based, and percentage tip pools with the FLSA Section 3(m)(2)(A) tip-credit rules built in. HotSchedules Tip Manager is a paid add-on at roughly $1.50-3/employee/mo. Toast Tips Manager is built into Toast Payroll and is free if you run Toast end-to-end — this is the biggest soft argument for going all-in on Toast at a single-location restaurant. Whichever tool you pick, model the pool offline first because mistakes here are wage-theft claims; NY Labor Law §196-d prohibits employers and managers from sharing in the pool, and tip credit only applies if you give written notice per NYS Hospitality Wage Order. Run the pool calculation through the system for 4 weeks alongside your old method before cutting over.

Realistically: ±10-15% on weekly sales after 6-8 weeks of historical data, ±20-30% on first opening month, and noticeably worse on weather days, holidays, and special events. The AI engine is just a moving average with day-of-week, weather, and event overlays — it cannot anticipate a viral TikTok or a subway shutdown. Treat the forecast as a starting point that your GM should override 2-3x per week; the value is that it gets you to a draft schedule in 20 minutes instead of 2 hours. The number to track is forecast variance: the gap between predicted sales and actual sales, reported weekly. If your variance is under 12% you're scheduling tighter labor; over 20% you're either over-staffing on slow days or running 86s on busy ones.

Allow swaps with auto-approval only when both employees hold the same role, neither hits overtime (NY 40-hour threshold for time-and-a-half), and the swap doesn't land within the 11-hour rest window between shifts that NY Labor Law treats as the practical safe harbor against spread-of-hours violations. 7shifts and HotSchedules both have a swap-rules engine — set those three guardrails and let the rest go. Without auto-approval, your manager spends 30-60 minutes a day rubber-stamping swaps; with the right rules, that drops to 5-10 minutes a week reviewing edge cases. The risk you're managing is overtime: NYS hospitality OT kicks in at 40 hrs/week and an unwatched swap can add a $200-400 OT line you didn't budget for.

Toast Payroll if you run Toast POS — the integration auto-pulls hours, tips, and declared cash tips, and runs $39/mo base plus $5/employee/mo. Gusto if you don't run Toast — $40/mo base plus $6/employee/mo on the Simple plan, with tip credit and NYS hospitality wage rules built in and the cleanest UX in the category. Paychex Flex and ADP Run are the two enterprise picks at roughly $100-200/mo base plus $10-15/employee/mo, plus setup fees of $200-500; pick them only if you have 50+ employees per location or need robust HR-compliance hand-holding. The single biggest implementation mistake is letting payroll run separately from POS and re-keying tips and hours every week — that's where the 3-7% error rate comes from and where wage-claim exposure lives.

NYS 2026 minimum wage in NYC is $16.50/hr; the tipped food-service worker minimum is $11.00/hr cash wage with a $5.50/hr tip credit, per NYS Hospitality Industry Wage Order (12 NYCRR Part 146). You must (1) give the employee written notice of the tip credit before applying it (Section 146-2.2), (2) confirm the employee's tips bring them to at least $16.50/hr each week, and (3) keep records for 6 years. If tipped time falls below 80% of hours worked in a week (the '80/20' practical rule still observed by NYDOL), you owe full minimum wage for that week. Configure your payroll system to flag any week where a server's tips + cash wage falls below $16.50/hr — Toast Payroll and Gusto both do this automatically once you tag the role as tipped.

NYS spread-of-hours (12 NYCRR §146-1.6) requires one extra hour of pay at the basic minimum wage ($16.50 in NYC for 2026) any time an employee's workday spans more than 10 hours from start of first shift to end of last shift, including unpaid breaks. This hits split-shift servers hardest — a server who works 11am-3pm lunch and 5pm-10pm dinner has a 11-hour spread and is owed the extra hour. Toast Payroll, Gusto, and ADP all auto-calculate spread-of-hours when you tag the location as NYC-covered and feed clock-in data from the POS. Audit one pay period manually before trusting the calc; missed spread-of-hours pay is one of the top-3 NYDOL wage claims against NYC restaurants and back-pay can run 6 years.

NYS Labor Law §191(1)(a) requires manual workers (which courts have held includes most BOH and many FOH roles in restaurants) to be paid weekly and not later than 7 days after the end of the workweek. After Vega v. CM & Associates Construction (2019) and the 2024-25 wave of NY appellate decisions, manual workers paid bi-weekly may sue for the late-payment liquidated damages even though they received the full wages — this is a live class-action target. The defensible posture for any NYC restaurant: run weekly payroll for hourly tipped/non-exempt staff, and bi-weekly or semi-monthly for salaried managers only. Toast Payroll, Gusto, and ADP all support weekly cycles at no extra fee per run; the only cost is your bookkeeper's time.

Push credit-card tips through payroll (next-day or 3-day direct deposit) — it eliminates the safe-keeping risk, gives you a clean wage record, and is what 70%+ of NYC FSR have moved to since 2022. Toast Tips Manager and 7shifts Tip Payouts both let you pay credit-card tips next-business-day via instant-pay rails (Branch, Wisely, Toast Pay Card) at $0-2/payout per employee. Cash tips you collect at the table from physical cash payments are still distributed in cash at end of shift, but the employee declares them on a tip declaration form that feeds payroll for tax purposes. The single biggest operational gain: your closing manager stops counting and securing $2-5K in tip cash every night, which removes a real theft and safety risk.

If you have 10+ employees in NYC and have been in business at least 2 years, the NYC Retirement Security for All Act requires you to either (a) sponsor a qualified retirement plan (401k, SIMPLE IRA, etc.) or (b) enroll employees in NY State's Secure Choice IRA — auto-enrolling at 3% of pay. The NY State program rolled out in 2024-25; enforcement and penalties are now active in 2026. Gusto, Toast Payroll, and Guideline (the leading 401k startup, partners with both) can set up a turnkey Safe Harbor 401k for $40-150/mo + $4-8/employee/mo, which is usually cheaper administratively than running NY Secure Choice once you scale past 15 employees. If you have 50+ employees you also pick up ACA employer mandate exposure; talk to a benefits broker before you cross 49.

Run your own online ordering — every operator who only relies on third-party marketplaces is paying 15-30% on every order to DoorDash/UberEats forever. Toast Online Ordering is included free on Core+ plans and has the cleanest POS integration; ChowNow charges $99-149/mo flat with $0/order commission (positioned as the anti-marketplace); Square Online is free with 2.6% + $0.10 per online order plus a small per-transaction fee. The realistic NYC mix at a healthy single-location restaurant is 50-60% dine-in, 20-25% first-party online (your website), 15-20% third-party marketplaces — drive customers from marketplaces to your own site by including a 10%-off card in every delivery bag. The savings: a $50 order placed direct nets you ~$48; on DoorDash it nets you ~$35-38.

Olo is enterprise-only — you should look at it once you're past 5-10 NYC units or when you're integrating ordering across multiple POS systems and need a single ordering layer. Olo (NYSE: OLO since 2021) runs the ordering for Shake Shack, Sweetgreen, sweetgreen, Wingstop, Five Guys, Chipotle, and most national chains; pricing is custom but realistically $500-2,000/mo per location plus per-order fees. The pitch is that Olo Rails connects you to every marketplace (DoorDash, UberEats, Grubhub, Postmates) through one menu and one inventory feed — so a single 86 instantly disappears from every channel. For a 1-3 unit operator the integration overhead and minimum spend kill the deal; stay on Toast Online or ChowNow until you need Olo's scale.

Yes if catering is more than 10% of revenue — a real catering ordering page needs lead-time enforcement (24-72 hour minimums), per-person pricing, delivery zone mapping, and corporate billing/PO support, which the standard online-ordering tools handle poorly. Toast has a Catering & Events module ($99/mo add-on) that handles this; ezCater (acquired by RSE Ventures, public via SPAC, NYSE: EZCT then went private 2025 at $1.6B) is the marketplace play with 15-18% commissions; CaterCow and Hungry are NYC-strong on the corporate side. For a single restaurant, start with Toast Catering at $99/mo plus a clearly labeled 'Catering' button on your homepage with a 24-hour minimum lead time. Catering orders average $15-50/pp at 5-15x the dine-in check, and the gross margin is typically higher than dine-in because labor is concentrated.

Only for high-volume bars, beer halls, food halls, and counter-service hybrids — never for full-service restaurants where the server is the experience. Toast Mobile Order & Pay and Square QR Order & Pay are both ~$25-50/mo add-ons; they cut labor cost 10-20% in the right concept and bump average ticket 5-15% because guests order more rounds without flagging a server. The risk in FSR is that tip percentage drops 200-400bps because guests perceive less service when they self-order. The right NYC use case: rooftop bars, beer gardens (Houston Hall, Standings), beach clubs, and any concept doing 200+ covers a service with table turnover under 60 minutes. For traditional sit-down service, QR ordering is a gimmick that costs you tips and guest experience.

Yes — a loyalty program disconnected from online ordering captures maybe 30% of eligible transactions; an integrated one captures 70-90%. Toast Online + Toast Loyalty are the simplest single-vendor stack at no extra integration cost; Square Online + Square Loyalty same story. If you run third-party loyalty (Punchh, Thanx, Como, Paytronix), make sure it has a native connector to your ordering platform — Olo integrates with all of them, ChowNow with Punchh and Thanx natively. The number to track is loyalty member capture rate at online checkout: target above 60%, and if you're under 40% the friction is too high (probably asking for too much info before they can pay). One concrete win: pre-fill the loyalty number based on email match instead of asking for a phone number again.

NYC's permanent delivery commission cap (Local Law 2021/52, codified at Admin Code §20-846) limits third-party delivery platforms to 15% of order value for delivery fees and 5% for marketing/all other fees, capped at 23% total per order, for any restaurant that doesn't expressly opt into higher fees. This was upheld in DoorDash v. NYC at the 2nd Circuit in October 2024, so the cap is settled law in 2026. Operators can voluntarily opt into higher 'marketing' tiers (DoorDash's 'Plus' at 25%, 'Premier' at 30%) for better placement in the app — most NYC operators run at the basic 15% tier and rely on direct ordering for the rest. The DCWP enforces; if a platform is charging you over 23% without your written opt-in, you can complain at NYC.gov/DCWP and seek refunds.

No — DoorDash acquired Caviar from Square in August 2019 for $410M, and Caviar now runs as a higher-end consumer brand inside the DoorDash backend. If you opt onto Caviar you're effectively getting a curated tier of DoorDash exposure — same merchant dashboard, same commissions structure (capped at 15% delivery + 5% marketing per NYC law), but listed alongside the higher-end NYC restaurants Caviar still curates. There's no separate contract and no separate fee schedule. The realistic value: Caviar still drives a slightly different demographic in Manhattan (older, higher-AOV) so opting in is essentially free upside if you're already on DoorDash. Don't pay extra for any 'Caviar tier' a sales rep tries to upsell — it doesn't exist as a distinct product anymore.

On a $40 menu-priced order at the NYC 15% delivery fee + 5% marketing cap (20% total), the platform takes $8 — so you receive $32 gross, before food cost. After 28-32% food cost ($11-13) and labor allocation ($6-8), you're netting $11-15 contribution margin on that order. That's roughly half of what the same order nets you at dine-in or first-party online. The arithmetic is why every NYC operator needs to run a marketplace-pricing strategy: most raise menu prices on marketplaces 10-20% above dine-in (which is permitted as long as you don't violate the platform's price-parity policy — DoorDash dropped price-parity requirements in 2021 settlement). If you're not using marketplace markups you're losing $4-8 per order in unrecovered margin.

Yes if you take orders from 3+ marketplaces — running 4 separate tablets at the host stand is the #1 cause of missed orders, late acceptance, and 1-star reviews. Otter (now Tryotter, ~$50-200/mo per location), Cuboh (acquired by Grubhub 2022), and Chowly are the three NYC-active aggregators; all three pull DoorDash, UberEats, Grubhub, Postmates into a single tablet and inject them into your POS as one ticket. Toast launched its own first-party Order Hub in 2023 which is included free on Core+ — for Toast restaurants that's the obvious pick. The other operational win: aggregators let you 86 an item or pause a channel from one screen instead of four, which is the difference between recovering gracefully from a fryer fire and getting destroyed in reviews.

All three major platforms (DoorDash, UberEats, Grubhub) auto-refund customers for missing items, wrong items, and late deliveries — and historically passed 100% of those refunds through to the restaurant. Since 2023-24, DoorDash and UberEats both let you dispute individual refunds from the merchant dashboard and report 'error rates' that, if low, qualify you for refund protection programs ('Drive Errors' on DoorDash, 'Trusted Partner' on UberEats). The threshold is roughly under 5% error rate to qualify for any meaningful protection. The single biggest operational fix is a checklist tape printed at expo for every delivery order — cuts missing-item refunds by 60-80%. Track refund rate weekly: NYC FSR delivery refund rate over 6% means you have a packing problem that's costing you $400-2,000/mo at modest order volume.

Yes — DoorDash Drive and UberEats Direct let you keep first-party orders (placed on your website or by phone) but use the marketplace's driver fleet to deliver, paying $7-11 per delivery instead of the 15-20% commission on a marketplace order. For a $40 order this is a much better deal: you keep the 15-20% commission AND you collect customer data instead of the platform owning the relationship. Toast On-Demand Delivery, Olo Dispatch, and Relay all wrap multiple driver networks (DoorDash Drive, UberEats Direct, Postmates, Roadie) into a single API call. The catch: drivers are not your employees, you don't pick the time, and on a busy Friday night you can wait 25-40 minutes for a driver to accept — set realistic delivery time expectations on your site (45-65 min) so reviews don't hit you.

Yes, surprisingly — Grubhub still has 15-20% NYC delivery share in 2026, mostly from older Manhattan office workers and Seamless legacy users who never moved off the brand. Grubhub was acquired by Just Eat Takeaway for $7.3B in June 2021, then sold to Wonder (Marc Lore) for $650M in November 2024 — a >90% writedown — and is now operated as a delivery layer underneath Wonder's broader food platform. NYC operators should stay on Grubhub for the residual order volume but should not invest in any premium ad placements; the demand pool isn't growing. Grubhub commissions sit at the same 15% delivery + 5% marketing NYC cap as DoorDash and UberEats. Cancel any 'Grubhub Plus' marketing spend over the basic tier — it's a worse ROI than the same dollars on DoorDash sponsored placements.

Run digital gift cards through your POS (Toast Gift Cards, Square eGift, Lightspeed Gift) — included free or near-free on the standard plans, with no per-card cost beyond standard processing. For physical gift cards order through Plastek, Plasmart, or PaperWorks (typical cost $0.40-1.20/card minimum order 500), branded with your logo and a magnetic stripe or QR code that ties to your POS gift card module. NYS Abandoned Property Law §1315 says gift certificates issued in NY do not expire and cannot be charged a dormancy fee until 5+ years of inactivity — set your POS rules accordingly. Gift cards are essentially free working capital: NYC FSR breakage (cards never redeemed) runs 8-15%, which means $1,000 of holiday gift card sales typically nets you $80-150 in pure margin from breakage alone.

Points-based ($1 spent = 1 point; 100 points = $10 reward) wins for full-service NYC restaurants because it scales with check size and rewards higher spenders proportionally. Visit-based (buy 9, get the 10th free) wins only for true habit-formers — coffee shops, juice bars, the burrito spot near the office. Toast Loyalty and Square Loyalty both offer both models at $25-99/mo, included on higher-tier plans. The realistic NYC FSR target: 25-40% of guest checks tied to a loyalty member, with member checks averaging 15-25% higher than non-member checks. Punchh and Thanx (mid-market loyalty platforms at $200-1,000/mo) add segmentation, lifecycle automations, and SMS push but you really only need them once you're past 5K active members or multiple locations.

Capture emails at three points: online ordering checkout (mandatory field), reservation booking (Resy/OpenTable both export to email automation), and gift-card purchase. SMS opt-in is harder under the TCPA — require explicit double opt-in with a clear keyword reply ('YES'), or you're exposed to $500-1,500/text statutory damages. Toast Marketing, Mailchimp, and Klaviyo are the three common platforms — Klaviyo is the gold standard for NYC restaurants over 2K subscribers at $20-150/mo depending on list size, with prebuilt restaurant flows (welcome series, win-back, birthday). Realistic 90-day target for a single NYC restaurant: 1,500-3,000 email subscribers, 500-1,000 SMS opt-ins, 25-35% email open rate, 2-5% click-to-order conversion on a weekly send.

The two common attacks are (1) staff voiding a sale and pocketing the cash equivalent in gift card balance, and (2) third parties social-engineering balance lookups by phone. Lock down by requiring manager PIN approval for any gift card balance reversal or transfer in your POS, and disabling phone balance lookup or requiring last-4 of the original purchaser's credit card. Toast and Square both log every gift card transaction with employee ID — pull a weekly gift card activity report and flag any clerk who issues 3+ gift cards/week without a corresponding sales transaction. The other lever: a $500-1,000 daily gift card sale ceiling per terminal, raised only with manager override. NYC FSR gift card fraud loss runs 0.3-0.8% of gift card sales — anything over 1% means you have an internal problem.

Use a dedicated connector — Shogo ($30-50/mo per location), Sourcery (now Xtra Chef, owned by Toast as of 2022), or QuickBooks' native Toast/Square apps — to push daily sales summaries into QuickBooks/Xero as journal entries. The connector maps POS sales categories (food, NA bev, alcohol, gift card, comps, voids) into QB chart of accounts on a daily cadence, eliminating ~6 hours/wk of bookkeeper data entry per location. Do not push individual transactions into QB; you want one journal entry per day with the proper categorization. The validation step: tie out POS gross sales to bank deposits weekly using a 3-way match (POS report, merchant processor statement, bank deposit), and reconcile any variance over $50 same-day.

MarginEdge ($330/mo flat, owned by Compeat/Restaurant365 as of 2022 acquisition) and Restaurant365 ($289-459/mo per location) are the two NYC standards for inventory + invoice automation tied to POS sales depletion. Both pull POS sales nightly, depletion-cost against your recipes, and produce daily theoretical-vs-actual food cost reports. Toast xtraCHEF and Toast Inventory (Toast's first-party stack) are cheaper at $99-249/mo if you're already on Toast and don't need full back-office accounting. The unsexy reality: inventory tools only work if someone actually counts inventory — most NYC operators do a weekly count (sundays close) and a monthly hard count. Without counts the system reports 'theoretical' food cost only, which is useful as a directional metric but doesn't catch theft, over-portioning, or spoilage.

Yes — Toast Developer API (toasttab.com/developers) is publicly documented, REST-based, with OAuth 2.0 and webhooks for orders, menu, employees, and labor. Most of the public endpoints are read-only or limited write; full menu and order write access requires a Toast Partner Program agreement (free to apply, but Toast vets the use case). Square's API is broader and easier to get into for free; Lightspeed K-Series has a robust REST API as well. Aloha and legacy Micros require expensive certifications and are effectively closed to small developers. If you're building a one-off integration (digital signage that pulls daily specials, etc.) Toast and Square both work with a few weeks of dev time; if you want to resell anything to other operators, you'll need the partner agreement.

iPad-based POS (Square, Lightspeed, TouchBistro) costs less upfront ($799 for an iPad Pro vs $1,299 for a Toast Flex terminal) but has a meaningful reliability tradeoff in a kitchen environment — iPads are not rated for heat, grease, or being dropped on tile. The 24-month total cost of ownership tends to favor purpose-built terminals (Toast Flex, Clover Station Solo, Lightspeed Stand) once you account for replacement rate. Toast and Clover units come with a 1-year limited warranty plus paid extended; iPads from Apple come with the standard 1-year warranty with AppleCare+ extending to 3 years at $79-149 per device. The defensible setup for a busy NYC restaurant: purpose-built terminals at the host stand and bar, iPad handhelds for table-side ordering (Toast Go, Lightspeed Tableside).

Validate the HMAC signature on every incoming webhook — Toast, Square, and Stripe all sign payloads with HMAC-SHA256 using a secret you provision in the dev portal. Reject any request without a valid signature, log the source IP, and rate-limit aggressively (a real POS will not send more than ~50 webhooks/sec even at peak). Use HTTPS only and rotate the webhook secret every 90 days. Store the raw payload + signature for at least 30 days for debugging and replay protection. The single most common security mistake: trusting the order_id or amount in the webhook payload without re-fetching the resource via authenticated API call before acting on it — always fetch and verify before mutating state in your downstream system.

Epson TM-m30III (thermal receipt, ~$300) for the host stand and bar, Epson TM-U220B (impact dot-matrix, ~$350) for any kitchen printer that touches heat — thermal printers fade and jam in kitchen humidity. Star Micronics is the credible alternative at similar pricing. Both Epson and Star ship with USB, ethernet, and Bluetooth interfaces; you want ethernet for any printer in a fixed location, never WiFi. Budget ~$25-40 per printer for thermal paper rolls per month at moderate volume, and keep a 6-month supply on hand because your POS reseller will mark up paper 2-3x. Skip 'mobile' printers for the kitchen — the impact paper handles grease and heat where thermal does not.

Spec a dedicated business-grade fiber line (Verizon Fios Business or Spectrum Business, 200/200 Mbps minimum, ~$100-200/mo NYC) plus a 4G/5G failover (Verizon, T-Mobile, Cradlepoint device, ~$50-80/mo) so the POS doesn't go down when ConEd or Spectrum has an outage. Run a managed router (Ubiquiti UniFi Dream Machine Pro, ~$379, or Meraki MX67 ~$595 + license) with separate VLANs for POS, BOH, BOH IoT, guest WiFi, and security cameras — a flat network with everyone on one SSID is the #1 NYC SMB security weakness. Hardwire every POS terminal and KDS screen with Cat6; reserve WiFi for handhelds and guest. Budget $1,500-3,500 one-time for the router, switches, cabling, and APs at install.

Yes, because NYC Local Law 89 of 2020 (Admin Code §20-845) prohibits restaurants from refusing to accept cash payment from in-person customers, with civil penalties up to $1,500/violation enforced by DCWP. You need a cash drawer at every register that takes in-person orders (host stand, bar, counter). The standard APG Vasario or M-S Cash Drawer Argox at $150-250/each connects to most receipt printers via RJ12; the cash drawer opens when the printer fires a 'kick' command on a cash transaction. Cash now runs 5-15% of NYC FSR sales and 10-25% of bar sales — meaningful but not dominant. Skip cash drawers at table-side handhelds; servers should drop cash at the closeout station to a single locked drop box.

Yes — EMV chip and contactless tap (NFC) are mandatory in practice because as of October 2015 (chip) and 2017 (tap), the chargeback liability shifts to the restaurant for any fraudulent magstripe-only transaction. Toast Flex, Square Terminal, and Clover Station Solo all ship with EMV + NFC + magstripe; pin-on-glass (typing PIN on the touchscreen rather than a separate keypad) is approved under PCI MPoC standard and supported by all three. Apple Pay and Google Pay tap-to-pay run through the same NFC reader at no extra cost and account for 25-40% of NYC card transactions in 2026. Skip any reader without contactless — your guests will judge it harshly and you'll get held up at the table while a server runs the card to the back.

Keep at least one fully-imaged spare terminal in the office, two backup receipt printers, a spool of thermal paper covering 30 days, two spare iPad handhelds (if you use them), a spare router, and one backup card reader. The single largest service-killer is a host-stand printer dying at 7pm Saturday with no replacement — cost of failure is $2K-5K in slowed turn rate over the night vs $300 for a spare printer. Also keep a powered USB-C and Lightning charging station, 10 spare receipt rolls, 4 spare thermal rolls for KDS printers, and Cat6 patch cables in 1ft, 3ft, 7ft. Document all serial numbers and Toast/Square device IDs in a single spreadsheet so swapping in the spare takes 5 minutes, not 45.

Yes — a single $200-400 UPS (APC SMT1500 or CyberPower CP1500) on your router/switch/modem stack keeps the POS, KDS, and credit-card processing alive through the typical 5-30 minute NYC power blip without losing in-flight transactions. Run a separate smaller UPS at the host stand to keep the printer alive during the same window; line cooks have flashlights, but a printer that won't fire when power flickers means tickets stop flowing. Total UPS budget for a typical NYC restaurant: $600-1,200 across 2-3 units, replaced every 3-4 years as batteries degrade. The other failure mode UPS protects against: dirty power that fries terminals — Toast Flex and iPad units both have cheap power supplies that don't tolerate brownouts well.

Toast Standard install: 4-6 weeks from signed contract to live, assuming you have menu PDFs and floor plan ready on day one. Square: 3-7 days because most setup is self-serve. Lightspeed: 4-8 weeks. Aloha/Simphony: 12-20 weeks because of the menu programming, on-site install, and training overhead. The single biggest delay is menu engineering — every modifier, sub-modifier, and price point has to be entered exactly, and most operators underestimate this by 2-3x. Block off two full days of an experienced manager's time to build the menu in the POS, and another full day for floor plan and table mapping. Schedule the cutover for a Tuesday after lunch service so you have 4 days of low-volume shifts to debug before the first weekend.

Build the menu in a structured Google Sheet first — columns for category, item, price, all modifier groups (with min/max selection rules), modifier prices, tax category, and printer routing. Toast, Square, Lightspeed, and Olo all import from CSV/Excel; manual entry of a 100-item menu takes 8-15 hours, importing from a clean sheet takes 2-3 hours plus QA. Have the executive chef and GM both review the menu in the POS test environment before go-live and run 5-10 mock orders covering every modifier path — this catches roughly 80% of opening-week ringup errors. Don't skip the tax category mapping: NYC sales tax (8.875%) applies to prepared food and most non-alcoholic beverages, but bottled water and most groceries are exempt — getting this wrong creates either NYS sales tax exposure or guest disputes.

Run two 90-minute group training sessions (one FOH, one BOH) the week before go-live, then keep the old system live in 'view-only' mode for the first 7 days while the new one runs primary. All four major POS vendors (Toast, Square, Lightspeed, Clover) offer training-mode environments where staff can practice on real workflows without affecting actual sales — use these for a week before go-live. Pair each new server with a 'POS buddy' (often a manager or stronger server) for the first 3 shifts. Budget $400-1,500 in additional manager hours for the cutover week, and expect a 5-15% slowdown in service throughput for week 1 that recovers fully by week 3. Filming a 5-minute screen-record walkthrough for each major workflow (open table, add item, modifier, split check, close out) cuts training time roughly in half.

Plan for 4-8 weeks of overlap and budget $5-15K in implementation costs (hardware, install, training time, dual subscriptions), plus a 10-20% revenue dip in the first 2 weeks of cutover at a busy single-location restaurant. The most painful elements are gift card balances (must be exported and manually loaded into the new system, with strong reconciliation), loyalty member data (often loses transaction history), reservation history (Resy/OpenTable/SevenRooms migrate cleanly via API but the new POS guest profiles start fresh), and historical reporting (you'll keep the old POS subscription running 3-12 months as a read-only archive). The right window: a slow January or August Tuesday — never the week before a Friday-Saturday peak. The wrong reason to switch: minor feature gaps; you can fix most with a $50/mo add-on cheaper than a re-platform.

Day-of-soft-open checklist: every menu item ringable with all modifier paths tested, sales tax mapping verified against an NYS-tax-rate calculator, gift card module live and tested with a $10 test sale + redemption, EMV + tap card readers tested with at least one chip card and one Apple Pay tap, KDS routing tested for every station, employee logins and PINs created for every scheduled staff member, manager-only voids/comps configured with PIN, time clock pulling from the POS to the scheduling tool, online ordering page live with hours and lead times correct, gift cards able to be sold both in-store and online. Run a full mock service Tuesday afternoon with all staff before Friday soft open. The single most common opening-night failure is a printer routing rule that sends apps to the dishwasher pit instead of expo because someone tested with one terminal and not the live floor layout.

Toast and Square have 'offline mode' — terminals continue to take orders and swipe cards (storing them locally) as long as the terminal itself has power, even if the cloud connection drops. Confirm offline mode is enabled on every terminal at install (it's default-on at Toast, default-on at Square, default-off at Lightspeed K-Series — check yours). For full network outages, fall back to a paper ticket book and a Square Reader on a personal cellphone hotspot — keep two such backup readers in the manager's drawer. Document a written outage SOP: who calls support (Toast 617-682-0225, Square 1-855-700-6000), who tells the kitchen, who walks the floor and apologizes. The single biggest fail mode is staff trying to wait it out for 30 minutes hoping the system comes back; they should be writing tickets within 5 minutes of confirmed outage.

NYC FSR effective rate (total processing fees / total card volume) typically lands at 2.7-3.2% in 2026 — anything above 3.3% means you're being overcharged. Toast Payments quotes 2.49-2.99% + $0.15; Square quotes 2.6% + $0.10 in-person and 2.9% + $0.30 online; Stripe and most online processors hit 2.9% + $0.30. Add ~10-25bps for AmEx (which now generally processes at parity but historically charged more), 20-50bps for international cards, and assessment/network fees of 13-20bps that are passed through transparently in interchange-plus pricing but bundled (and often padded) in flat-rate pricing. The single best rate-reduction lever above $3M/yr in card volume is moving from Toast/Square's flat rate to interchange-plus pricing through a third-party processor (Worldpay, Stripe Restaurant, Heartland) — typical savings of 30-60bps, worth $9-18K/yr at $3M volume.

Toast and Square offer month-to-month software at the standard rate, but Toast Payments hardware financing comes with a 36-month commitment if you take the 0% financing — early termination triggers full hardware payoff plus a $250-500 processor early-termination fee per location. Push for a 12-month max term on any new POS deal, walk away from any 36-month + auto-renewal clause that requires 60-90 day written cancellation notice. Clover/Fiserv and ADP/Aloha contracts often hide $295-495/terminal early-termination fees in the fine print — read sections 4-7 of any merchant services agreement before signing. The realistic posture: pay the higher month-to-month rate ($69-165/mo) rather than save $20/mo on a 3-year commitment that locks you into hardware and processing rates that will be uncompetitive in 18 months.

If you take any card payment you fall under PCI DSS — practically, you need to complete an annual SAQ (Self-Assessment Questionnaire) and a quarterly external network scan. Toast, Square, Clover, and Lightspeed all bundle PCI compliance support with their POS subscription and most operators in NYC complete the SAQ B-IP or SAQ P2PE in 30-90 minutes once a year. The realistic risk is a non-compliance fee from your processor — typically $25-100/mo if you let the SAQ lapse — plus the catastrophic $50-90 per compromised card if you suffer a breach (Target paid $18.5M to settle 47-state AGs in 2017; small operator settlements typically $5K-50K). The defensible posture: never store full card numbers anywhere outside the POS, never email or text card numbers between staff, and segment guest WiFi from POS network on separate VLANs.

Yes, with strict rules — NY GBL §518 was enjoined as unconstitutional after Expressions Hair Design v. Schneiderman (2017) and a follow-on 2020 settlement, so a NY merchant can now surcharge credit cards (not debit) up to the actual processing cost or a 4% cap (whichever is less), as long as the price is posted as the cash price with the surcharge clearly disclosed before the swipe. NYS DOS issued formal guidance in 2024 reinforcing the disclosure requirement: posted at the entrance, on the menu, and at the point-of-sale. Most NYC operators in 2026 either (a) bake processing into menu pricing, (b) offer a 'cash discount' (legally cleaner than a surcharge), or (c) add a 3-3.5% credit card surcharge with conspicuous disclosure. Surcharging will save you 2.5-3.5% on every credit-card sale but can dent guest sentiment — survey your guests before deploying.

Toast: send written cancellation to support@toasttab.com and request the 'cancellation packet' — they will quote you the remaining hardware financing balance plus any unamortized implementation credit. Negotiate by citing competing quotes; Toast will sometimes waive the early-term fee if you commit to keep using Toast Payroll or another product. Clover/Fiserv: harder, because the contract is with the local ISO (independent sales organization) reseller, not Fiserv directly — find your reseller name on the merchant statement, send certified-mail cancellation 90 days before any auto-renewal date, and document the request multiple ways. The leverage: file a BBB complaint and a NYS DFS complaint if the reseller resists; both have moved Clover ISOs to release operators in past disputes. Budget $1-5K to pay out remaining hardware financing on most early cancellations.

A merchant reserve (rolling reserve, capped reserve, or upfront reserve) is what processors do to protect themselves against your future chargebacks — usually 2-10% of monthly card volume held for 90-180 days before release. New restaurants with no processing history routinely face a 90-day rolling reserve at 5-10%, which can tie up $30K-200K of working capital you didn't budget for. Ask the processor in writing what triggers reserve release (typically 6-12 months of clean processing) and negotiate the reserve down at signing — Toast, Square, and Stripe all rarely reserve below $500K/mo volume; smaller ISO-resold deals often do. If you've been processing 12+ months with under 0.5% chargeback rate, file a written reserve-reduction request quarterly until they release. The reserve is your money — track it as a current asset on the balance sheet.

EMV chip and contactless tap transactions where the cardholder physically presented the card carry zero merchant liability for fraudulent-use chargebacks — you win these automatically by submitting the EMV transaction record. The chargebacks you actually fight are 'service not received,' 'incorrect amount,' and 'duplicate charge' — you have 5-7 days from notice (varies by network) to respond through the processor portal with itemized receipt, signed credit card slip if available, server name, and any communication history. Win rate on properly defended NYC FSR chargebacks: 40-60%. Track your chargeback rate weekly: anything over 1% triggers Visa/Mastercard monitoring programs at $5K+ in fines, and over 2% can get you placed on the MATCH list (the processor blacklist) which makes finding a new processor extremely expensive. The easiest prevention is a clear merchant DBA name on the card statement that matches your storefront sign.

Read the merchant statement line by line every month for: PCI non-compliance fee ($20-100/mo if SAQ lapses), monthly minimum fee ($25-50/mo if you don't hit a volume floor), batch fee ($0.10-0.30 per daily settlement), gateway fee ($10-30/mo for online), AVS fee ($0.05/transaction), regulatory compliance fee ($5-15/mo, often pure markup), statement fee ($5-15/mo paper statement), and the 'IRS reporting fee' ($3-7/mo, also pure markup). On a Toast/Square statement these are largely absent or itemized clearly; on Clover/Fiserv and old-school ISO resellers they can total $80-200/mo of pure markup that has nothing to do with your actual processing cost. Demand a fee schedule in writing at signing and re-audit every 6 months — processors quietly add fees in 'rate review' notices on page 4 of your statement.

Yes — every processor (Toast, Square, Stripe, Clover, all of them) charges your card processing rate (2.6-2.99% + $0.10-0.15) on the full ticket including tip, because the card network treats the tip as part of the authorized amount. On a $100 dinner with a $20 tip, you pay processing on $120 — roughly $3.50, of which ~$0.60 is on the tip portion. Across a year at $1.5M card volume with a 20% tip ratio, you eat ~$8-9K in processing fees on tips you immediately pass through to staff. There is no clean way to avoid this short of running tip-out off cards (e.g., cash tipping only), which is operationally bad. The defensible response: factor tip-share processing into your effective labor cost when modeling pricing, and make sure your menu pricing accounts for it.

Yes — Toast, Square, Lightspeed, and most modern POS vendors will export your customer database (email, phone, transaction history, loyalty balances, gift card balances) on request, usually as CSV or via API, at no charge. Pull this export at least quarterly even when you're not planning to leave; treat it as a backup against vendor-side data loss or service termination. The grey area is 'derived data' (segments, scores, RFM analysis the vendor built on top of your raw data) — vendors typically argue they don't owe you that, and they're right under most TOS. Read your contract's data clause before signing: any modern POS should explicitly state you own your customer data; if it doesn't, walk. The CCPA/CPRA-style data-portability movement (and likely future NYS analog) is making this more enforceable each year.

Buy direct in nearly every case — resellers (independent sales organizations, or ISOs) add a 25-100bps markup on processing rates plus monthly fees that direct customers don't pay. The pitch from local resellers is 'we'll be on-site to fix things,' but Toast and Square both have NYC-based field service that responds in 4-24 hours, and most issues are resolved by phone or remote support. The exception: large multi-unit operators (10+ locations) who can negotiate volume pricing through a strategic ISO partner sometimes get genuinely better rates than direct, but those deals are rare and require >$5M/yr card volume. The other exception is Clover, which is essentially only sold through ISOs (and that's part of the problem). For 1-5 unit NYC operators, buy Toast or Square direct and skip the reseller pitch.

Toast publicly targets 99.9% cloud uptime (about 8.7 hours of downtime per year) but the standard merchant agreement does NOT include a financial SLA — there's no service credit if they miss it. Square is similar: 99.9% target, no financial penalty for missing. Oracle Simphony in its enterprise tier offers a contractual 99.9-99.99% SLA with service credits for misses, which is part of why luxury hotels still pay for it. The practical implication: Toast and Square outages do happen (notable ones: Toast 4-hour outage Feb 2024, Square 3-hour outage September 2023) and you have zero contractual recourse beyond reputational pressure. This is why offline mode and a backup payment path matter — your remediation comes from your own preparation, not the vendor's promise.

Most ISO-resold POS contracts (Clover, Aloha, legacy Micros) auto-renew for 12 or 24 months unless you send written cancellation notice 60-90 days before the renewal date — miss the window and you're locked in for another full term, with the same early-termination fees as before. NY GBL §5-903 limits auto-renewal of certain consumer service contracts but commercial contracts generally fall outside that protection. Diary the cancellation deadline 6 months before contract expiration and again 4 months out. Toast and Square month-to-month contracts don't have this trap. The defensive move at signing: strike the auto-renewal clause entirely and replace with 'shall convert to month-to-month at end of term' — most resellers will agree if you push.

If a card breach is traced to your environment (compromised terminal, malware on a back-office PC, misconfigured network), you face PCI fines of $5K-100K from your processor, $50-90/card in network reissuance fees, plus state-by-state breach notification requirements (NY SHIELD Act requires notification of all affected NY residents and the AG/DOS/State Police if 500+ NY residents are affected). NYC FSR breach settlement examples in 2024-25 averaged $15K-150K all-in for small operators. Cyber liability insurance specifically covering PCI fines and breach response runs $1.5K-5K/yr for $1M coverage at a single-location restaurant — buy it. The single biggest risk reduction: get rid of any back-office PC running Windows that touches your POS network, segment guest WiFi onto its own VLAN, and require manager passwords (not staff PINs) for any settings changes. Toast and Square cloud architecture limits your direct breach exposure compared to legacy on-prem systems.